ZurichOpenRepositoryand Archive Year: 2012 · 2020. 8. 27. · Mutationsstudien zeigten, dass...
Transcript of ZurichOpenRepositoryand Archive Year: 2012 · 2020. 8. 27. · Mutationsstudien zeigten, dass...
-
Zurich Open Repository andArchiveUniversity of ZurichMain LibraryStrickhofstrasse 39CH-8057 Zurichwww.zora.uzh.ch
Year: 2012
The role pf Gaf-1b in intracellular trafficking and autophagy
Diep, Tu-My
Posted at the Zurich Open Repository and Archive, University of ZurichZORA URL: https://doi.org/10.5167/uzh-68814DissertationPublished Version
Originally published at:Diep, Tu-My. The role pf Gaf-1b in intracellular trafficking and autophagy. 2012, University of Zurich,Faculty of Science.
-
The Role of Gaf-1b in Intracellular Trafficking and
Autophagy
Dissertation
zur
Erlangung der naturwissenschaftlichen Doktorwürde
(Dr. sc. nat)
vorgelegt der
Mathematisch-naturwissenschaftlichen Fakultät
der
Universität Zürich
von
Tu-My Diep
von
Bern
Promotionskomitee
Prof. Dr. Peter Sonderegger (Vorsitz)
Prof. Dr. Jack Rohrer
Dr. Uwe Konietzko
Zürich 2012
-
CONTENTS
Contents
Summary ................................................................................................................... 1
Zusammenfassung ................................................................................................... 3
Abbreviations ............................................................................................................ 5
Publications .............................................................................................................. 9
1. Introduction ......................................................................................................... 11
1.1 Intracellular Transport ................................................................................................................. 11
1.2 Secretory Pathway ...................................................................................................................... 13
1.3 Endocytic Pathways .................................................................................................................... 15
1.4 Endocytic Pathways in polarized cells ........................................................................................ 18
1.5 Autophagy ................................................................................................................................... 20
1.6 SNAREs ...................................................................................................................................... 23
1.7 Rab proteins ................................................................................................................................ 25
1.8 Rab11 and Rab11-interacting proteins ....................................................................................... 27
1.9 The calsyntenin family ................................................................................................................. 29
1.10 GABARAP-likes and MAP1LC3 family ..................................................................................... 32
1.11 Aims .......................................................................................................................................... 35
2. Results ................................................................................................................. 37
2.1 Gaf-1b is a novel interaction partner of calsyntenin-1 ................................................................ 37
2.2 Study of the interaction of Gaf-1b and calsyntenin-1 .................................................................. 39 2.2.1 Calsyntenin-1 interacts specifically with Gaf-1b ................................................................................... 39
2.2.2 The C2 domain of Gaf-1b is involved in the interaction with calsyntenin-1 .......................................... 40
2.2.3 The acidic stretch of calsyntenin-1 is crucial for the interaction with Gaf-1b ........................................ 42
2.2.4 The formation of Gaf-1b/Rip11 complex disrupts the interaction of calsyntenin-1 and Gaf-1b ............ 44
2.3. Gaf-1b and calsyntenin-1 are proteins of recycling endosomes ................................................ 45
2.3.1 Gaf-1b and Rip11 exclusively interacts with GTP-Rab11 .................................................................... 45
2.3.2 Calsyntenin-1 interacts with the recycling-endosomal protein Rab11 through Gaf-1b ......................... 48
2.3.3 Gaf-1b localizes with endosomal recycling and TGN marker in neurons ............................................. 49
2.4 Gaf-1b is transported in calsyntenin-1 vesicles of the recycling-endosomal pathway................ 50 2.4.1 Gaf-1b and Rip11 are selectively associated with the recycling-endosomal type of calsyntenin-1
transport packages ....................................................................................................................................... 50
-
CONTENTS
2.4.2 Gaf-1b is coupled to the calsyntenin-1/syntaxin 13 complex. .............................................................. 51
2.4.3 Gaf-1b is transported in calsyntenin-1 organelles devoid of APP ........................................................ 53
2.4.4 Gaf-1b interacts simultaneously with calsyntenin-1 and kinesin-light chain 1 ...................................... 54
2.5 Gaf-1b trafficking is disrupted upon blockage of calsyntenin-1 and KLC1 TGN exit .................. 56
2.6 Gaf-1b and receptor recycling ..................................................................................................... 58
2.7 Overexpression of Gaf-1b and calsyntenin-1 affects the morphology of HeLa cells .................. 60 2.7.1 Gaf-1b knock-down induces stress fibre formation .............................................................................. 60
2.8 Gaf-1b is necessary for cell viability ............................................................................................ 62
2.9 GABARAPL1 and GABARAPL2 interact with Gaf-1b ................................................................. 64 2.9.1 GABARAPL1 and GABRAPL2 are novel interaction partners of Gaf-1b .............................................. 64
2.9.2 Localization of the binding domain of Gaf-1b for the interaction with GABARAPL2 ............................. 66
2.9.3 Gaf-1b colocalizes with GABARARPL1 and GABARAPL2 .................................................................. 67
2.10 Gaf-1b is involved in the autophagic pathway .......................................................................... 69
2.10.1 Gaf-1b associates with membranes of the autophagosomal pathway ............................................... 69
2.10.2 Gaf-1b accumulates in perinuclear aggregates upon starvation ........................................................ 70
2.10.3 Gaf-1b knock-down induces autophagosomes .................................................................................. 72
3. Discussion .......................................................................................................... 73
3.1 Gaf-1b interacts with calsyntenin-1 ............................................................................................. 73
3.2 Gaf-1b is associated with the TGN ............................................................................................. 76
3.3 Gaf-1b is transported in TGN-derived calsyntenin-1 organelles along axons ............................ 77
3.4 Gaf-1b is associated with the endosomal recycling compartment and interacts with endosomal
recycling markers .............................................................................................................................. 77
3.5 Gaf-1b has a minor effect on receptor recycling ......................................................................... 81
3.6 Gaf-1b is involved in the autophagic pathway ............................................................................ 82
3.7 Gaf-1b is essential for cell viability .............................................................................................. 84
4. Material and Methods ......................................................................................... 87
4.1 Antibodies and Reagents ............................................................................................................ 87 4.1.1 Affinity-purified antibodies .................................................................................................................... 87 4.1.2 Purchased antibodies .......................................................................................................................... 87 4.1.3 Provided antibodies ............................................................................................................................. 88 4.1.4 Reagents ............................................................................................................................................. 88
4.2 Constructs ................................................................................................................................... 89 4.3 siRNA and shRNA constructs ..................................................................................................... 91
4.3.1 siRNA ................................................................................................................................................... 91 4.3.2 shRNA ................................................................................................................................................. 91
4.4 Molecular cloning ........................................................................................................................ 93 4.4.1 PCR and RT-PCR ................................................................................................................................ 93 4.4.2 Purification of PCR products ................................................................................................................ 93 4.4.3 Restriction digest ................................................................................................................................. 93
-
CONTENTS
4.4.4 Agarose gel electrophoresis ................................................................................................................ 94 4.4.5 Gel extraction ....................................................................................................................................... 94 4.4.6 Ligation ................................................................................................................................................ 94 4.4.7 Preparation of competent E.coli cells ................................................................................................... 94 4.4.8 Transformation of recombinant plasmid into competent E.coli XL1-blue or DH5α ............................... 95 4.4.9 Transformation of competent E.coli BL-21 ........................................................................................... 95 4.4.10 Isolation and purification of DNA from E.coli cells .............................................................................. 95
4.5 Cell biological methods ............................................................................................................... 96 4.5.1 Heterologous cell culture ..................................................................................................................... 96 4.5.2 Neuron cell culture ............................................................................................................................... 96 4.5.3 Transfection of adherent heterologous cells with Polyethylenimine or LipfectamineTM 2000 ............... 96 4.5.4 Transfection of siRNA in adherent heterologous cells ......................................................................... 97 4.5.5 Amaxa nucleofection ............................................................................................................................ 97 4.5.6 Immunocytochemistry .......................................................................................................................... 98 4.5.7 Starvation experiment .......................................................................................................................... 99
4.6 Biochemical methods ................................................................................................................ 100 4.6.1 SDS-PAGE ........................................................................................................................................ 100 4.6.2 Coomassie brilliant blue staining ....................................................................................................... 100 4.6.3 Silver staining .................................................................................................................................... 100 4.6.4 Immunoblotting .................................................................................................................................. 100 4.6.7 GST Pull-down ................................................................................................................................... 101 4.6.8 Immunoprecipitation ........................................................................................................................... 101 4.6.9 Subcellular fractionation ..................................................................................................................... 101 4.6.10 Immunoprecipitation from solubilized brain membrane fraction ....................................................... 102 4.6.11 Immunoisolation of vesicular organelles .......................................................................................... 102 4.6.12 Biotinylation experiments ................................................................................................................. 102 4.6.13 Yeast two-hybrid screen .................................................................................................................. 103
4.7 Statistics .................................................................................................................................... 103
5. References ........................................................................................................ 105
Curriculum Vitae ................................................................................................... 115
Acknowledgment .................................................................................................. 117
-
SUMMARY
1
Summary
Regulated endocytosis and recycling is an essential process for cells to maintain their function
and retain the composition of the plasma membrane. Recycling is involved in several cellular
processes such as cell adhesion and junction formation, cell migration, cytokinesis, morphogenesis,
cell fusion, cell polarity, signal transduction, nutrient uptake, learning and memory. Another important
process is autophagy. This plays an essential role as a physiological system for cellular homeostasis.
Many neurodegenerative diseases, such as Alzheimer’s disease, Parkinson’s disease or Huntington’s disease, are linked to defective autophagy due to the inability to clear aggregates of mutated toxic
proteins. The understanding and treatment of such diseases relies on detailed knowledge of the
molecular machinery that governs and controls autophagy.
Calsyntenin-1 is a neuronal type-1 transmembrane protein of the cadherin superfamily found
at postsynaptic sites in the adult mouse brain. We recently showed that amyloid precursor protein
(APP) exits the trans-golgi network (TGN) in association with calsyntenin-1 in tubulovesicular
organelles in a kinesin-1-dependent manner and that these vesicles are transported anterogradely
along axons. Additionally, we found axons contained at least two distinct, non-overlapping calsyntenin-
1-containing transport packages, one is characterized by the presence of APP and early-endosomal
markers, the other with recycling-endosomal markers. Organelle immunoisolation and biochemical
approaches indicate that calsyntenin-1 provide a protective mechanism for axonal transport of APP.
Recently, Gaf-1b, a splice variant of the Rab11-interacting protein Rip11, was identified as a novel
interaction partner of calsyntenin-1. Here we characterized the interaction and the physiological
function of Gaf-1b and calsyntenin-1. Mutational studies showed that Gaf-1b specifically interacts with
the C-terminus of calsyntenin-1 through its C2 domain. We suggested that Gaf-1b is C-terminally
bound to membranes through the association with GTP-Rab11, and that the splice insert of Gaf-1b
was necessary to facilitate a conformation by which the N-terminus of Gaf-1b can interact with
calsyntenin-1. Overexpression of calsyntenin-1 and light chains (KLC1) of kinesin-1 impaired overall
ER and Golgi structure and aggregated Gaf-1b in TGN subdomains. In addition, immunoisolation of
calsyntenin-1- and Gaf-1b-containing organelles from mouse brains and colocalization studies in
cultured neurons indicate that Gaf-1b/Rab11 is recruited to the TGN, leaves the TGN in a calsyntenin-
1-dependent manner, and is involved in the tethering of TGN-derived tubules to recycling endosomes.
Biochemical approaches showed a direct interaction of Gaf-1b with Rab11 and a potential interaction
between calsyntenin-1 and syntaxin 13, suggesting that calsyntenin-1 links SNARE complexes
containing syntaxin 13 to Gaf-1b, which in turn recruits the small GTPase Rab11. This complex may
play a role in the accurate tethering and fusing of recycling vesicles with other organelles or with the
plasma membrane. Interestingly, high expression of Gaf-1b and calsyntenin-1 in HeLa cells resulted in
an alteration of the cell shape and in extensive outgrowth of the plasma membrane. Thus, we
assumed that calsyntenin-1 and Gaf-1b/Rab11 have a specific function in regulating plasma
membrane expansion, cytoskeleton rearrangement and/or receptor recycling in axonal growth cones
or spine growth at postsynaptic sites. We further demonstrated that Gaf-1b interacts specifically and
directly with the two MAP1LC3A/Atg8 family members GABARAPL1 and GABARAPL2. Members of
the MAP1LC3/Atg8 family are associated with the process of autophagy. In HeLa cells we found that
Gaf-1b associates with subcompartments of the autophagosomal pathway. Induction of autophagy by
starvation mediates perinuclear aggregates positive for Gaf-1b and the autophagosomal marker LC3.
Downregulation of Gaf-1b affected the level of autophagosomes in HeLa cells. Furthermore,
-
SUMMARY
2
pharmacological studies indicated that Gaf-1b may play a role in the downstream steps of the
autophagosomes formation.
In summary, our data indicated a direct relation between endosomal organelles in the
autophagic pathway. Gaf-1b may regulate the recruitment of multivesicular bodies (MVBs) masked
with Rab11 and GABARAPL1 and GABARAPL2 to autophagosomal membrane to induce the fusion of
MVBs with lysosomes.
-
ZUSAMMENFASSUNG
3
Zusammenfassung
Regulierte Endozytose und Recycling von Membranbestandteilen ist ein essentieller Prozess
für Zellen, um ihre Funktion aufrecht zu erhalten. Recycling nimmt Teil an verschiedenen zelluläre
Prozessen, wie die Bildung von Zellkontakten und Zelladhäsion, Zellwanderung, Zellteilung,
Morphogenese, Zellfusion, Zellpolarität, Signaltransduktion und Nahrungsaufnahme. Ein anderer
wichtiger zellulärer Prozess ist die Autophagie, welche eine essentielle Rolle für die zelluläre
Homöostase spielt. Gestörte Autophagie kann zu diversen neurodegenerativen Erkrankungen, wie
z.B. Alzheimer, Parkinson und Huntington führen, da die Zelle Aggregate von mutierten toxischen
Proteinen nicht entfernen kann. Das Verständnis der molekularen Grundlagen dieser Erkrankungen
und deren Behandlung sind daher abhängig von einer detaillierten Aufklärung der molekularen
Maschinerie, welche Autophagie vermittelt und kontrolliert.
Calsyntenin-1 ist ein neuronales Typ-1-Transmembranprotein der Cadherin-Superfamilie, das
an der postsynaptischen Seite im adulten Nervensystem vorkommt. Wir haben kürzlich gezeigt, dass
Amyloid-Precursor-Protein zusammen mit Calsyntenin-1 das Trans-Golgi-Netzwerk in tubulär-
vesikulären Organellen mittels eines Kinesin-1-abhängigen Prozesses verlässt. Diese Vesikel werden
anterograd entlang der Axone transportiert. Zusätzlich haben wir herausgefunden, dass Axone
mindestens zwei verschiedene, nicht überlappende Calsyntenin-1-enthaltende Transportvesikel
besitzen. Die einen sind durch die Anwesenheit von Amyloid-Precursor-Protein und early-
endosomalen Markern charakterisiert und die anderen durch die Marker der Recyclingendosomen.
Immunoisolierung von Organellen und biochemische Studien weisen darauf hin, dass Calsyntenin-1
einen schützenden Mechanismus für den axonalen Transport von Amyloid-Precursor-Protein bietet.
Neulich wurde Gaf-1b, eine Splice-Variante des Rab11-bindenden Proteins Rip11, als neuer
Interaktionspartner von Calsyntenin-1 identifiziert. Wir haben hier die Interaktion und physiologische
Funktion von Gaf-1b und Calsyntenin-1 charakterisiert. Mutationsstudien zeigten, dass Gaf-1b über
die C2-Domäne spezifisch mit dem Carboxy-Terminus von Calsyntenin-1 interagiert. Wir nehmen an,
dass der Carboxy-Terminus von Gaf-1b über die Wechselwirkung mit GTP-Rab11 an die
Plasmamembran bindet und der Splice-Einschub von Gaf-1b notwendig ist, damit Gaf-1b eine
Konformation annehmen kann, in welcher der Amino-Terminus von Gaf-1b mit Calsyntenin-1
interagieren kann. Überexpression von Calsyntenin-1 und die leichte Untereinheit von Kinesin-1
veränderte die Struktur des endoplasmatischen Retikulums und des Golgi-Apparates und reicherte
Gaf-1b in Subdomänen des Trans-Golgi-Netzwerkes an. Zusätzlich zeigten die Immunoisolierung von
Calsyntenin-1- und Gaf-1b-enthaltenden Organellen aus Maushirn und Co-Lokalisations-Experimente
in Neuronenkulturen, dass Gaf-1b/Rab11 ins Trans-Golgi-Netzwerk rekrutiert werden, und dieses
dann in einer Calsyntenin-1-abhängigen Art verlassen. In der Folge sind Gaf-1b/Rab11 am Anbinden
von tubulären Organellen, die aus dem Trans-Golgi-Netzwerk entsprangen, an Recyclingendosomen
beteiligt. Biochemische Experimente zeigten eine direkte Interaktion von Gaf-1b mit Rab11. Wir
fanden auch Hinweise für eine Interaktion zwischen Calsyntenin-1 mit Syntaxin 13. Solche
Interaktionen könnten in Anbindungs- und Fusionsprozessen von Recyclingvesikeln mit anderen
Organellen oder der Plasmamembran eine Rolle spielen. Interessanterweise konnten wir in HeLa
Zellen extreme Auswüchse der Plasmamembran beobachten, wenn Gaf-1b und Calsyntenin-1 stark
überexprimiert wurden. Daraus folgerten wir, dass Calsyntenin-1 und Gaf-1b/Rab11 eine Funktion in
der Regulation von Plasmamembranexpansion, Cytoskelett-Reorganisation und/oder Rezeptor-
Recycling in axonalen Wachstumskegeln oder dendritischen Dornen haben. Weiter haben wir gezeigt,
-
ZUSAMMENFASSUNG
4
dass Gaf-1b spezifisch und direkt an zwei Mitglieder der MAP1LC3A/Atg8-Familie, nämlich
GABARAPL1 und GABARAPL2, bindet. Mitglieder der MAP1LC3A/Atg8-Familie sind in Prozesse der
Autophagie involviert. In HeLa Zellen ist Gaf-1b mit Subkompartimenten des autophagosomalen Pfads
assoziiert. Induktion von Autophagie durch Nährstoffentzug förderte die Bildung von perinukleären
Aggregaten, die Gaf-1b und den autophagosomalen Marker LC3 enthalten. Verminderte Expression
von Gaf-1b beeinflusste die Zahl der Autophagosomen in HeLa Zellen. Weitere pharmakologische
Studien weisen darauf hin, dass Gaf-1b eine Rolle in weiteren Schritten der Autophagosomenbildung
spielt. Zusammengefasst deuten unsere Daten auf einen direkten Zusammenhang zwischen
endosomalen Organellen des autophagosomalen Pfads hin. Gaf-1b könnte dabei die Rekrutierung
von multivesikulären Strukturen, die mit Rab11 dekoriert sind, und GABARAPL1/GABARAPL2 an
autophagosomale Membranen regulieren und dann die Fusion von multivesikulären Strukturen und
Lysosomen induzieren.
-
ABBREVIATIONS
5
Abbreviations
aa amino acid
ADAM a disintergrin and metalloproteinase
AICD APP intracellular domain
AMPA alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid
APP amyloid β-protein precursor ATP adenosintriphosphate
BACE β-site APP cleavage enzyme CAM cell adhesion molecule
CCP clathrin-coated pit
CCV clathrin-coated vesicles
CD-M6PR cation-dependent mannose-6-phosphat receptor
CNS central nervous system
COPI/II coat protein complex I/II
Cst calsyntenins
CTF C-terminal fragment
Cy3 cyanine 3
Cy5 cyanine 5
DAPT N-[N-(3,5-difluorophenacetyl)-L-analyl]-S-phenylglycin t-butyl ester
DIV days in vitro
DMEM Dulbecco’s modified eagle medium DMSO dimethyl sulfoxide
cDNA complementary deoxyribonucleic acid
pcDNA plasmid deoxyribonucleic acid
DTT DL-dithiothreitol
ed ectodomain
EDTA ethylene diamine tetracetic acid
EEA1 early endosome antigen 1
EGFP enhanced green fluorescence protein
EGFR epidermal growth factor receptor
EHD3 Eps 15 homology domain-containing protein 3
ER endoplasmatic reticulum
Erb type 1 transmembrane receptor tyrosine kinase
ERC endosomal recycling compartment
ERM ezrin-radixin-moesin
ERGIC ER-Golgi intermediate compartment
EtBr ethidium bromide
EtOH ethanol
FCS fetal calf serum
FIPs Rab-family interacting proteins
FITC fluorescein isothiocyanate
fl full-length
GABA Gamma-aminobutyric acid
-
ABBREVIATIONS
6
GABARAP GABAA receptor associated protein
GABARAPL 1/2 GABAA receptor associated protein ½
Gaf-1b γ-SNAP-associated factor 1b GAP GTPase-activating protein
GATE-16 Golgi-associated ATPase Enhancer of 16 kDa
GDI GDP dissociation inhibitor
GDP guanidinediphosphate
GEC1 glandular epithelial cells
GEF GDP/GTP exchange factor
GFP green fluorescence protein
GM130 Golgi martrix protein of 130kD
GST Glutathione S-Transferase
GTP guanidinetriphosphate
HA hemagglutinin
HBSS++ Hanks Balanced Salt Solution (1 mM MgCl2, 1 mM CaCl2, and 1% vol/vol
human serum albumin)
HRP horseradish peroxidase
IC ER-Golgi intermediate compartment
ICC immunocytochemistry
IgG immunoglobulin G
IP immunoprecipitation
KBS KLC-1 binding segment
kD kilo Dalton
KHC kinesin-heavy chain
KLC kinesin-light chain
LAMP lyosome associated membrane protein
LC light chain
LDL low density lipoprotein
LTP long-term potentiation
LB Luria-Bertani broth
MAP microtubule-associated protein
MDCK Madin-Darby canine kidney
mRFP monomeric red fluorescence protein
MS mass spectormetry
MTOC microtubule-organizing centre
NA numerical aperture
NMDA N-methyl-d-aspartate
NSF N-ethylmaleimide-sensitive factor
NTP nucleoside triphosphate
dNTP deoxynucleoside triphosphate
PA phosphatidic acid
PBS phosphate-buffered saline
PCR polymerase chain reaction
PEI polyethylenimine
PFA paraformaldehyde
-
ABBREVIATIONS
7
PLL poly-L-lysine
PM plasma membrane
PMA phorbol 12-myristate 13-acetate
PNS postnuclear supernatant
POD peroxidase
PSD postsynaptic density
PI(3)P phosphatidylinositol (3)-phosphate
PtdIns(3,4,5)P3 phosphatidylinositol-(3,4,5)-triphosphate
PVDF polyvinylidene fluoride
RBD Rab11/25 binding domain
RCP Rab11 coupling protein
REP Rab escort protein
RFP red fluorescence protein
Rip11 Rab11 interacting protein
RNA ribonucleic acid
RNAi RNA interference
mRNA messenger ribonucleic acid
rpm revolutions per minute
RT room temperature
RT-PCR reverse transcriptase poly chain reaction
SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis
SE sorting endosome
siRNA small interfering RNA
SNAP soluble NSF association protein
SNARE SNAP receptor
STEM surface-connected tubules entering macrophage
Stx syntaxin
TAE Tris-acetate EDTA (ethylene diamine tetracetic acid)
TBST Tris-buffer saline Tween
Tf transferrin
TfR transferrin receptor
TGN trans-golgi network
TM transmembrane
TPR tetratricopeptide repeat
UV ultra violet
VAMP vesicle associated membrane protein
WB Western blot
wt wild-type
-
8
-
PUBLICATIONS
9
Publications The following publications have been obtained during my PhD thesis:
Original articles:
05/2012 Steuble, M., Diep, T. M., Schatzle, P., Ludwig, A., Tagaya, M., Kunz, B. and
Sonderegger, P. (2012). Calsyntenin-1 vesicles shelter APP from proteolytic
processing during anterograde axonal transport. Biology Open 000, 1-14.
08/2010 Steuble, M., Gerrits, B., Ludwig, A., Mateos, J. M., Diep, T. M., Tagaya, M.,
Stephan, A., Schatzle, P., Kunz, B., Streit, P. et al. (2010). Molecular
characterization of a trafficking organelle: dissecting the axonal paths of
calsyntenin-1 transport vesicles. Proteomics 10, 3775-88.
01/2009 Ludwig, A., Blume, J., Diep, T. M., Yuan, J., Mateos, J. M., Leuthauser, K.,
Steuble, M., Streit, P. and Sonderegger, P. (2009). Calsyntenins mediate TGN
exit of APP in a kinesin-1-dependent manner. Traffic 10, 572-89.
Selected Meeting Abstracts:
10/2012 Diep, T. M., Steuble, M., Schatzle, P., Ludwig, A., Tagaya, M., Kunz, B. and Sonderegger, P. (2012). Calsyntenin-1 vesicles shelter APP from proteolytic processing during anterograde axonal transport. Society for Neuroscience, New Orleans. Abstract submitted and accepted.
06/2012 Steuble, M., Diep, T. M., Schatzle, P., Ludwig, A., Tagaya, M., and Sonderegger, P. (2012). Calsyntenin-1 vesicles shelter APP from proteolytic processing during anterograde axonal transport. ZNZ and NCCR Neuro Joint Symposium, Zurich.
05/2011 Diep, T. M., Steuble, M. and Sonderegger, P. (2011). The role of Gaf-1b in vesicular transport of calsyntenin-1 in neurons. ZNZ PhD Retreat, Valens.
02/2011 Steuble, M., Diep, T. M., Kunz, B., and Sonderegger, P. (2011). A role of calsyntenin-1 for sheltered APP transport along axons. NCCR Neuro (National Center of Comptence in Research Neural Plasticity and Repair), Warth.
03/2009 Diep, T. M., Ludwig, A., Blume, J., Yuan, J., Mateos, J. M., Leuthauser, K., Steuble, M., Streit, P. and Sonderegger, P. (2009). Calsyntenins mediate TGN exit of APP in a kinesin-1-dependent manner. Swiss Society of Neuroscience, Fribourg.
03/2008 Diep, T. M., Steuble, M., Ludwig, A. and Sonderegger, P. (2008). APP and calsyntenin-1: differently regulated proteolytic processing and trafficking in axonal endosomal subcompartments. NCCR Neuro (National Center of Comptence in Research Neural Plasticity and Repair), Berlingen.
-
10
-
INTRODUCTION
11
1. Introduction
1.1 Intracellular Transport Eukaryotic cells have developed an elaborate endomembrane system that compartmentalizes
biochemical pathways and biosynthetic processes. Each organelle is characterized by a specific
protein composition that needs to be sustained. Intracellular transport plays an important role in the
maintenance of the specific protein composition of distinct compartments. There are two major
pathways in the intracellular trafficking: the secretory pathway and endocytic pathway. The secretory
pathway involves the transport of molecules from the endoplasmic reticulum (ER), through the Golgi
apparatus, toward the cell periphery. The distinct membrane-bound compartments are interconnected
by vesicular traffic for secretion of newly synthesized proteins, carbohydrates and lipids. On their way
to the extracellular space, newly synthesized secretory proteins pass through several membrane-
enclosed organelles until they reach their maturation. This process includes the translation and folding
at the endoplasmatic reticulum, the package into COPII coated vesicles for transport to the Golgi
complex where they undergo modifications, and, as a last step, leaving the trans-Golgi network (TGN)
through the secretory granules (Figure 1.1) [2]. The endocytic pathway involves the internalization of
molecules from the plasma membrane or the extracellular space into endosomes. The fate of proteins
is determined by sorting processes in endosomes. This pathway comprises of several distinct
endocytic organelles, including sorting endosomes for protein sorting, late endosomes and lysosomes
for protein degradation. Additionally, the endocytic recycling describes the pathway through which
endocytosed proteins (such as receptors) are recycled back to the plasma membrane through the
recycling endosomes. Endocytic and exocytic organelles communicate constantly with each other
through the aid of the transport vesicles. These vesicles continually bud from one membrane and fuse
with another membrane, thereby transporting membrane components and soluble cargo [3].
Figure 1.1: Intracellular transport pathways.
The compartments of secretory and endocytic pathways are shown. The location of the different coats is indicated: COPII
(blue), COPI (green) and clathrin (red). ER, endoplasmatic reticulum; ERES, ER exit site; ERGIC, ER-Golgi intermediated
compartment; MTOC, microtubule-organizing center; TGN, trans-Golgi network (adapted from [4]).
-
INTRODUCTION
12
Intracellular trafficking, cell movement and subcellular localization of organelles are dependent
on the cytoskeleton. Microfilaments, intermediate filaments and microtubule are the three main types
of cytoskeletal filament. The actin filaments are the thinnest filaments of the cytoskeleton. They are
composed of two coiled strands of chains of actin subunits [5]. Their orientation is more random
throughout the cell, but there are bundles of actin filaments that are mainly concentrated directly
beneath the plasma membrane for the maintenance of cellular shape and formation of cytoplasmic
protrusions, such as pseudopodia or microvilli. Actin filaments are polar structures with a barbed end
and a pointed end [5]. Actin filaments are involved in the participation of cell-to-cell or cell-to-matrix
junctions. They also function in vesicle endocytosis, cell migration, cytokinesis and, along with myosin,
in muscular contraction. During vesicle biogenesis actin filament dynamics promote membrane
deformation, such as creation of tubular bud on nascent transport carriers during post-Golgi transport,
or the production of spherical carriers during endocytosis [6]. Intermediate filaments are more stable
than actin filaments and, like actin filaments, they participate in the maintenance of cell-shape and in
the cell-to-cell and cell-to-matrix junctions. Intermediate filaments are also involved in the organization
of internal three-dimensional structure of the cell and anchoring organelles. Microtubules are hollow
cylindrical filaments comprising of protofilaments. Protofilaments are polymers of alpha and beta
tubulin. Tubulin polymerizes end-to-end with an alpha-subunit of one tubulin dimer contacting the
beta-subunit of the next [7]. Therefore, in a protofilament one end will have the alpha-subunit exposed
and the other end will have the beta-subunit exposed, referred to as minus and plus end, respectively.
This polarity is an important feature of the microtubule structure. Microtubules are involved in many
cellular processes, including vesicular transport, determination of cell shape, cytoplasm organization,
mitosis, and cytokinesis [7]. Microtubules are nucleated and they are organized by microtubule-
organizing centres (MTOCs), such as centrosomes. The plus end of the microtubule is pointed toward
the cell periphery and the minus end is pointed toward the centrosome [8]. Microtubule-associated
proteins (MAPs) regulate the microtubule dynamics. Membranous carriers are transported within the
cell through motor proteins moving along microtubules in a defined direction. Kinesin and dynein are
the major motor protein of the microtubule. Kinesin moves toward the plus end of the microtubule,
therefore transports cargos from the centre of the cell towards the periphery. This form of transport is
referred to as anterograde transport. In contrast, dynein normally moves toward the minus end of the
microtubule. The movement of transport packages from the periphery toward the centrosome is
defined as retrograde transport. Thus, motor proteins link the cargo to the microtubule by interacting
directly with a transmembrane protein, or indirectly, via adaptor proteins.
Organelles communicate with each other through transport of vesicles. The size and form of
the vesicle varies from small and round (40 nm - 100 nm), to large and irregular tubulovesicular (≥200 nm). The transfer of cargo between compartments involves different proteins and contains distinct
steps, namely vesicle budding, protein sorting, vesicle targeting and vesicle fusion (Figure 1.2). Firstly,
coat proteins are assembled to induce membrane curvature and to promote budding of new vesicle
from a donor organelle [9, 10]. Then, cargo proteins are selectively sorted into the forming vesicles,
while resident proteins are retained in the donor compartment. The cargo selection is then mediated
by coat proteins recognizing sorting signals present in the cytosolic domains of transmembrane cargo
proteins. In a next step, the formed vesicle dissociates from the donor membrane by a mechanism
called scission. Subsequently, vesicles lose their coat on the way to the acceptor membrane.
Targeting and fusion to a specific acceptor compartment is strictly regulated by specific proteins.
Tethering is mediated by Rab proteins and SNAREs are involved in membrane fusion.
-
INTRODUCTION
13
Figure 1.2: Vesicular transport.
1) Coat proteins are assembled from the cytosol to initiate budding. 2) COPI, COPII or clathrin mediate the budding of transport
carriers and the uptake of cargo into transport vesicles. 3) The newly formed vesicle dissociates from the donor membrane with
different coats using different mechanisms for the scission. 4) The vesicle loses its coat on the way to their destination. 5) The
vesicle is transported to the acceptor membrane. The vesicle is guided by the cytoskeleton and is tethered to the acceptor
compartment by Rab GTPases and tethering factors. 6) Docking is mediated by SNARE complex formation where v-SNARE
and t-SNARE assemble into a four-helix bundle. 7) The assembled four-helix bundle drives the fusion of the vesicle with its
target compartment. The cargo is subsequently depleted into the acceptor compartment (adapted from [3]).
1.2 Secretory Pathway
The secretory pathway involves the delivery of various transmembrane and secretory proteins
to their target location and is essential for the cellular function. The secretory pathway comprises of
structurally distinct membrane-bound compartments and transport intermediates trafficking between
these compartments. The compartments of the secretory pathway include the ER, the ER exit sites
(ERES), the ER-Golgi intermediate compartment (ERGIC), the Golgi complex, and the trans-Golgi
network (TGN) [4].
The first station of the secretory pathway is the endoplasmatic reticulum, the site of protein
synthesis. The ER originally evolved through the invagination of the plasma membrane. Therefore, the
internal lumen of the ER is topically equivalent to the outside of the cell [11]. Its immense reticular
membranous structure is contiguous with the nuclear membrane. Almost the entire cytoplasm is
captured by the ER, and the ER consists of more than half of the total membrane in eukaryotic cells
[12]. The ER is involved in different important functions such as protein translation, folding and
modification, lipid synthesis, calcium homeostasis, and production and storage of glycogen, steroids
and other macromolecules. The ER can be distinguished from the nuclear envelope, and from smooth
(ribosome-free) and rough (ribosome-coated) membranes. Transmembrane or secreted proteins are
initially translated at ribosomes in the cytosol, and they enter the ER by the recognition of a signal
peptide found at the N-terminus of the protein. The signal recognition particle (SRP) recognizes the
signal peptide and forms a ribosome/nascent protein/SRP complex that binds to the ER membrane
through the SRP receptor [13]. At the ER membrane the nascent protein chain enters the ER lumen
through a proteinaceous pore in the membrane, namely the Sec61 translocon [11]. The translation
-
INTRODUCTION
14
proceeds and the newly synthesized proteins are released in the lumen of the ER, where protein
chaperons and lectins assure proper protein folding. Posttranslational modifications may take place
during this process. The nascent secretory proteins leave the ER at the ER exit sites (ERES). ERES
are specialized areas of the smooth ER that form buds on the nuclear envelope, and form
tubulovesicular networks in the periphery of the cell [14]. The ERES are coated with COPII coat
components. Soluble cargo proteins are sorted to the ERES by binding with cargo receptors that
belong to the ERGIC-53 family [15], the p24 family [16, 17] or the Erv family [18]. The lumen-exposed
domain of the receptors binds to the cargo proteins, while the cytoplasmic domain interacts with COPII
components [4]. COPII coat assembly starts with the recruitment of the small GTPase Sar1-GTP and
two sub-complexes consisting of Sec23/Sec24 and Sec13/Sec31 [14]. After budding, the COPII
coated transport vesicles are transported to the next compartment in the secretory pathway, a network
of tubulovesicular clusters, also known as ER-Golgi intermediate compartment (ERGIC) [19]. The
ERGIC is the major sorting station where ER residing proteins are recycled back to the ER via COPI-
mediated retrograde transport, and secretory cargos are delivered to the cis-Golgi [20] by forming a
new cis-Golgi or by fusing with an already existing cis-Golgi [19]. In the ER-to-Golgi transport the
microtubules play an important role because they associate with the ERGIC and mediate dynein-
dependent anterograde transport to the Golgi [21] and kinesin-dependent retrograde transport [22].
The Golgi apparatus is a multifunctional organelle next to the nucleus and closely associated
with the microtubule-organizing centre [23]. The Golgi in mammals is composed of a single ribbon of
stacked cisternae with a defining cis-to-trans asymmetry [24]. It has been shown in neurons that the
Golgi complex also adapts a ribbon-like network to a perinuclear position, but there is also Golgi
outposts observed in dendrites. These Golgi outposts represent many highly localized stations of the
secretory pathway [25]. Overall, the Golgi apparatus is the central station in the secretory pathway.
The Golgi apparatus is the place where post-translational modifications of secretory and
transmembrane proteins take place, such as glycosylation, sulfation and proteolytic processing. These
modifications in the Golgi are crucial for protein localization, stability and specificity of interactions [26].
Additionally, protein sorting is a major function of this organelle. Cargos approaching the Golgi from
the ER, fuse at the cis-side to form the cis-Golgi network (CGN), and are then transported and further
processed through the Golgi, to reach their final functional form at the trans-Golgi network (TGN). For
the traverse of proteins through the Golgi apparatus, several models have been proposed, such as the
anterograde vesicular transport and the cisternal maturation model. The anterograde vesicular
transport suggests that each individual cisterna is stable and immobile while cargo is transported
anterogradely in COPI coated vesicles that bud from one cisterna and fuse with the next one [20]. The
fact that COPI was shown to be responsible for retrograde transport, and that COPI vesicle only
transport small cargo molecules, render the anterograde vesicular transport model for intra-Golgi
transport questionable. In comparison, the cisternal maturation model is the most accepted model up
to date [27]. This cisternal maturation model states that cargo is kept within the same cisterna and
traverses from the cis- to the trans-face together with the cisterna. During this process, the cisterna
changes its composition by packaging of specific cisternal components, such as enzymes and lipids,
into COPI vesicles that move in the retrograde direction from the trans- to the cis-Golgi cisterna [28].
Secretory proteins reach the trans-Golgi network (TGN) after traversing the Golgi complex.
The TGN, a series of interconnected tubules that arise from three trans-Golgi cisternae [24], is the
major sorting centre of the secretory pathway. The size and structure of the TGN is very dynamic and
depends on the ratio of trafficking, the ratio of cargo input and output to the TGN along various
trafficking routes [26]. It was shown that trafficking blockage, induced by low temperature, results in an
-
INTRODUCTION
15
increased membrane surface area of the TGN [29]. The TGN does not only receive secretory proteins
from the trans-Golgi cisternae, but also recycling proteins from the endocytic pathway traverse through
the TGN [30, 31]. Therefore, the TGN is also an important node point for the communication between
the secretory and the endosomal system [1]. In the TGN, secretory proteins pass through their final
modification and then achieve their maturation through protease cleavage [20]. At the TGN, the newly
synthesized proteins are sorted to distinct pleiomorphic carriers that target different final destinations:
apical or basolateral membranes, recycling endosomes, early/sorting endosomes, late endosomes,
and specialized compartments, such as secretory granules in secretory cells (Figure 1.3) [1]. Vesicles
destined for the endosomal/lysosomal system are coated with clathrin coats. Only the tubules from the
last cisterna show clathrin-coated buds [32].
1.3 Endocytic Pathways
The endocytic pathways are used for internalization of molecules from the cell surface, such
as receptor-associated ligands, plasma membrane proteins, soluble molecules from the plasma
membrane, or the extracellular space and lipids. The removal of membrane from the cell surface is
balanced by endosomal recycling of the endocytosed proteins and lipids back to the plasma
membrane. The balance between endocytosis and recycling maintains the composition of the plasma
membrane, and is important for several cellular processes such as cell adhesion and junction
formation, cell migration, cytokinesis, cell polarity, signal transduction and nutrient uptake [33].
There are various mechanisms of endocytosis that can be distinguished on the level of
clathrin-dependent and clathrin-independent endocytosis [34]. In the last few years increased interest
in clathrin-independent endocytic pathways has arose. The best-known non-clathrin-coated pit
mechanism is caveolae formation. Caveolae are invaginations of the plasma membrane that are
coated with caveolin and the associated cavins [35, 36]. Other mechanisms of clathrin-independent
endocytosis include phagocytosis and pinocytosis. Phagocytosis refers to the uptake of large particles
like microorganisms or dead cells into phagosomes. This mode of internalization is induced by an
Figure 1.3: The TGN is a major sorting centre for
endocytic and exocytic pathways.
Secretory proteins traverse through the Golgi
complex to distinct domain of the TGN (I), where
they are transported to different destinations (1-5).
The TGN sorts proteins to apical plasma membrane
(1), the basolateral plasma membrane (2), recycling
endosome (3), early/sorting endosome (4), late
endosome (5) and specialized compartments, such
as secretory granules (6) in secretory cells. The
TGN also receives cargos from the endocytic
pathway (II-IV) and sends Golgi-resident proteins
back to the earlier Golgi complex (7). Only tubules
from the last cisterna are clathrin-coated. The ER is
in close contact with the tubules of the last two
cisternae for the exchange of lipids. Some apical or
basolateral proteins pass through the recycling
endosome before they reach the apical plasma
membrane (3, 3b) or basolateral plasma membrane
(3, 3a), respectively (adapted from [1]).
-
INTRODUCTION
16
actin-dependent process. This process usually occurs in specialized cells, such as macrophages, but
other cell types are also able to use phagocytosis [37]. Pinocytosis is the uptake of fluids into small
vesicles of about 100 nm. So far, the best characterized endocytic mechanism is receptor-mediated
endocytosis that involves the internalization of receptors and their ligands by clathrin-coated pits.
Clathrin is composed of three heavy and three light chains. They assemble to form three-legged cage-
like structures, referred to as a triskelion. The assembly of triskelia induces curvature in the membrane
[38]. Since triskelia do not directly interact with the membrane, a multiprotein complex, called adaptor
proteins (AP), is needed to recruit triskelia to the membrane [39]. AP-1 functions as clathrin-coated
vesicle (CCV) formation at the TGN, and AP-2 is responsible for the CCV formation at the plasma
membrane [40]. Alongside the recruitment of triskelia to the membrane and organization of CCP
formation, the APs also function as cargo docking proteins for molecules inside vesicles. To complete
the formation of clathrin-coated vesicles additional accessory proteins are needed. Epsins and
endophilin induce the bending of the membrane for vesicle budding [41], GTPase dynamin is required
for the fission of the vesicle from the donor membrane [42], and amphyphisin binds to endophilin,
clathrin, AP and dynamin [43].
Ligands and their plasma membrane receptors (e.g. TfR and LDLR) are internalized through
clathrin-coated pits, and the newly formed vesicles in the cell shed their clathrin coats. After the
clathrin coat disassembly the vesicles fuse with other newly formed vesicles, or with pre-existing
sorting endosomes. Sorting endosomes show a tubulovesicular structure with a light acidic luminal pH
of about 6.0, and they are located in the periphery of the cell [44]. Sorting endosomes allow fusion with
newly endocytosed vesicles for about 5-10 minutes. Afterwards, they start to move along microtubules
toward the cell centre, stop fusing with other endocytosed vesicles, and the lumen of the sorting
endosome becomes more acidic. Molecules in sorting endosomes are delivered to either one of the
four destinations, namely the TGN, the late endosome, the plasma membrane, or the endosomal
recycling compartment (ERC) (Figure 1.4). The low pH of sorting endosomes facilitates conformational
changes in proteins that can lead to the release of ligands from their receptors. The molecules
destined for degradation in lysosomes are first delivered to late endosomes. The sorting endosome
retains soluble ligands and membrane proteins in their lumen. Acidic hydrolases start to accumulate in
sorting endosomes, thus the sorting endosomes begin to mature to late endosomes, resulting in a pH
drop of about 0.5 units. Newly synthesized lysosomal proteins bound to mannose-6-phosphate
receptors (M6PR) are delivered from the TGN to late endosomes [45]. The luminal pH plays an
important role in the maturation of sorting endosomes. It was demonstrated that Bafilomycin A1, which
is a vacuolar H+-ATPase inhibitor, decreases the maturation process of sorting endosomes to late
endosomes and lysosomes [46, 47]. A goal of sorting signalling receptors to late endosomes is to
terminate receptor signalling, and to suppress cell responses to further signal input until new receptors
are synthesized [48]. Membrane receptors (e.g. EGFR, Erbs) are ubiquitylated in order to assign them
for downregulation. Ubiquitylated receptors are recognized by endosomal sorting complexes required
for transport (ESCRTs), that are also required for the biogenesis of endosomal multivesicular bodies
(MVB) [49]. The receptors are ubiquitylated at the cytoplasmic side. The ESCRTs recognize this
domain and induce inward-budding and the generation of intraluminal vesicles [50]. This separates the
cytoplasmic part of the receptor from the rest of the cell, and therefore suppresses their signalling
capability. Furthermore, MVBs fuse with lysosomes to degrade their luminal content.
While the sorting endosomes translocate to the centre of the cell before maturation to late
endosomes, most recycling proteins rapidly leave the sorting endosome by the budding of narrow-
diameter tubules [51, 52]. Sorting endosomes are separated in two distinct domains, namely the
-
INTRODUCTION
17
vacuolar domain and the network of tubules. Proteins that are retained in the vacuolar domain are
destined for degradation. Recycling proteins are accumulated in the network of tubules. Over 95% of
endocytosed proteins in sorting endosomes are removed before the maturation to late endosomes can
take place, and the t1/2 of endosome exit is 2 minutes or less [51, 52]. Thus, generating a tubular
geometry, where recycling proteins are translocated, is the first step of endocytic sorting (Figure 1.4;
right panel).
Figure 1.4: Model of receptor-mediated endocytic trafficking.
Left: Ligands bind at the plasma membrane to their receptors. The receptor-ligand complex is endocytosed mediated by clathrin
coats and transported to sorting endosomes. Vesicles can fuse with other newly formed vesicles or with pre-existing sorting
endosomes. Proteins destined for degradation are retained in sorting endosomes that mature into late endosomes and
lysosomes. Recycling proteins are recycled back to the plasma membrane either directly from the sorting endosomes or through
the endocytic recycling compartment (ERC). The ERC can receive proteins from the TGN and from endosomes. Proteins in
recycling endosomes are either sorted back to sorting endosomes, to the cell surface, or to the TGN. Some receptors and
molecules traffic between late endosomes and the TGN. Right: Recycling proteins are accumulated in the tubular domain of
sorting endosomes. From the tubular domain they are transported further to the endocytic recycling compartment. The vacuolar
domain matures to late endosomes and lysosomes and contains proteins destined for degradation. LDLR, Low-density-
lipoprotein receptor; LDL, low-density-lipoprotein; CL-MPR, cation-independent mannose-6-phophate receptor; TGN; trans-
Golgi network (adapted from [53]).
Recycling proteins are sorted back onto the cell surface through two main pathways: the fast
recycling route or the slow recycling route. One part of the recycling proteins in the sorting endosome
returns to the cell surface through the fast recycling pathway. In this pathway they are directly recycled
back to the plasma membrane from the sorting endosome. The other part takes the slow recycling
route through the endosomal recycling compartment [54]. Clathrin and its adaptor proteins have been
found on recycling endosomes, indicating that recycling endosomes may also form vesicles in a
clathrin-dependent manner [55, 56]. Therefore, recycling endosomes use similar mechanisms for
vesicle formation as do other transport intermediates forming compartments. The ERC (pH 6.4) is
slightly less acidic than the sorting endosomes (pH 6.0) and represents a heterogeneous and
tubulovesicular morphology. In some cells the ERC are distributed widely throughout the cytoplasm.
However, most of the ERC are typically localized juxtanuclear around the microtubule-organizing
centre next to the TGN, and they are associated with microtubules [55-57]. The ERC targets
-
INTRODUCTION
18
molecules to three different destinations: to the plasma membrane, to the sorting endosomes, or to the
trans-Golgi network. The transport from the ERC to either one of the destinations depends on the
generation of transport intermediates, vesicles or tubules. Different ERC residing proteins regulate the
transport from the ERC, such as Rab11 and the Eps15-homology-domain protein EHD1/Rme1
because the transport of various proteins to the TGN and to the plasma membrane is affected when
the activities of these proteins are altered [58]. Many plasma membrane proteins, such as Tac-
TGN38, which are destined for the TGN, pass through the ERC before reaching the TGN. It was
shown that TGN resident proteins can recycle several times between the plasma membrane and ERC
until they are delivered to the TGN [59]. The equilibrium of the protein distribution is maintained
because endocytic recycling is fast, compared to the exit rate from the Golgi. Many proteins, which are
not destined for the ERC, are still transported through the ERC on the way to their destination. Hence,
the ERC is an important sorting centre for molecules with distinct cellular locations.
1.4 Endocytic Pathways in polarized cells
Recycling pathways are essential for sorting molecules to their appropriate cellular locations,
and for the homeostasis of proper membrane and lipid distribution in secretory and endocytic
compartment. Many cell types display a polarized morphology by developing discrete functional
domains. Especially epithelial cells have distinct apical and basolateral domains defined by different
protein composition and separated by tight junctions. Neurons are highly polarized cells with two
structurally and functionally distinct domains, the axon and the dendrites [60-62]. Furthermore,
dendrites can be polarized into apical versus basolateral dendrites [63]. The establishment of neuronal
polarity begins early in the development of neurons. The maintenance and establishment of neuronal
polarity requires tightly coordinated regulation of the cytoskeleton and the cellular transport machinery
[64, 65]. The first step to establish a polarized neuron is the determination of a single axon, the axon
specification [66]. During axon specification, the cytoskeletal of one of the nascent neurites undergoes
prominent changes and the membrane composition is rearranged. The result is a neuritic process that
possesses different morphology and function from other neurites that are developing into dendrites
[67, 68]. The axonal growth cone is different from the dendritic growth cone in their protein
composition and development [69]. In the axon of mature neurons the microtubule has a polarized
orientation. The plus end points towards the distal part of the axon, away from the cell body. In
dendrites the microtubules possess mixed polarity [70]. The Rho family GTPases (Rho, Rac and
Cdc42) control cellular functions by regulating the actin cytoskeleton [71]. It was shown that the Rho
GTPases are important in establishing cell polarity [72] and influencing the neuronal morphology [73].
Reorientation of organelles of the secretory pathway during polarization is a result of a rearrangement
of the cytoskeleton. This creates a polarized secretory pathway with an orientation that supports
directional and selective membrane trafficking [74]. In epithelial cells, the Golgi apparatus is relocated
in the apical region of the cytoplasm, and protein delivery is directed towards the apical part of the
lateral plasma membrane [75]. In neurons, the Golgi is located in the soma prior to polarization, during
polarization it extends into larger dendrites [63].
To maintain the polarity of a cell, sustainment of the proper protein, lipid and cytoskeletal
composition within the distinct membrane subdomains is required. Several mechanisms have to be
coordinated to support polarization within cells. Barriers to control protein diffusion and stabilization of
protein complexes at the membrane by scaffolding proteins are necessary. Additionally, polarized
-
INTRODUCTION
19
trafficking through the endocytic and secretory pathways must occur [63]. Membrane trafficking in
polarized cells is best understood and described in epithelial cells. Proteins assigned for either the
apical or the basolateral membrane domain are directly sorted in the TGN into two distinct vesicle
species, in epithelial Madin-Darby canine kidney (MDCK) cells [76]. Another pathway, in which
proteins are sorted via transcytosis, was demonstrated for hepatocytes. In this pathway, membrane
proteins are first transported to the cell surface at the basolateral site. After internalization into
endosomes, proteins are sorted to the apical or basolateral membrane. Therefore, all the proteins are
first transported to the basolateral membrane, even if they are destined for the apical membrane [77].
It was suggested that basolateral and apical domains of MDCK cells correspond to the
somatodendritic and axonal domains of neurons, respectively [60]. The function of neurons is crucially
dependent on the morphological and functional differences between their axonal and dendritic
domains [60-62]. In addition to the distinction of axonal and dendritic domains, the neuronal polarity is
much more complex, as the dendritic and the axonal surface area are further polarized and
specialized into distinct subdomains [78]. Three different models have been proposed for the polarized
sorting of proteins to axonal and somatodendritic domains in neurons (Figure 1.5) [63]. The selective
pathway suggests that, at the TGN, axonal and dendritic proteins are sorted into separate transport
packages and selectively transported to axons or dendrites. Extensive studies showed that NgCAM
reach the axonal membrane through the selective delivery pathway [61, 79]. The FnIII domain at the
ectodomain of NgCAM is a sufficient sorting signal to assign NgCAM to the axon [61]. The selective
fusion route states that axonal and dendritic carriers are transported to both domains, but the fusion is
restricted to either one or the other target membrane [61]. The selective retention pathway postulates
that axonal proteins (such as VAMP2) are unselectively transported to axonal and somatodendritic
domains. At the dendritic membrane the proteins are not accumulated but rapidly endocytosed into
sorting/recycling endosomes and further sorted to the axon via transcytosis [61, 79, 80]. Thus,
endocytic recycling plays an important role in the establishment and maintenance of cellular polarity.
Figure 1.5: Three possible pathways for the polarized sorting of axonal and
dendritic plasma membrane proteins.
Three models have been proposed for polarized sorting of axonal and dendritic
plasma membrane proteins: (A) Selective delivery describes that axonal and
dendritic proteins are sorted directly at the TGN to their target destination. (B)
Selective fusion states that axonal proteins are delivered to both domains but
can fuse only with the axonal plasma membrane and not with the
somatodendritic membrane. (C) In selective retention, proteins are first
transported to the dendritic surface and after internalization, proteins are sorted
in endosomes to the axonal membrane (adapted from [63]).
-
INTRODUCTION
20
1.5 Autophagy
Autophagy is an evolutionary conserved and strictly regulated lysosomal degradation pathway
that is used by eukaryotes for the degradation of cytoplasmic components, such as long-living
proteins, entire organelles, and RNA. Therefore, autophagy plays an essential role as a physiological
system for cellular homeostasis. Autophagy occurs in virtually all cells at a low basal level. There are
three types of autophagy, namely macroautophagy, microautophagy and chaperone-mediated
autophagy [81, 82]. Microautophagy and chaperone-mediated autophagy take place directly at the
lysosomal membrane. Microautophagy involves direct engulfment of a small fraction of cytoplasm into
lysosomes by inward inversion of the lysosomal membrane. In Chaperone-mediated autophagy,
cytosolic proteins are unfolded by chaperone proteins and are translocated across lysosomal
membranes [83]. Among the three forms of autophagy, the most studied one is macroautophagy.
Macroautophagy (hereafter referred to as autophagy) involves the formation and expansion of a cup-
shaped structure, termed as isolation membrane or phagophore, that engulfs fractions of the
cytoplasm or organelles into a double membrane-bound vacuole, the autophagosome [84-86].
It has been demonstrated that autophagy plays an important role in vertebrate development.
Specific cytosolic rearrangements are required for proliferation, death and differentiation during
embryogenesis and postnatal development in mammals [86]. During embryogenesis there are
enhanced metabolisms or phases of silence in cells. Therefore, cells need to modify their organelle or
protein content quickly to rapidly adapt and respond to adverse conditions. Autophagy could fulfil
these requirements and be the dynamic tool to revive cells or modify their external appearance within
a few hours [86]. Autophagy is also involved in neurodegenerative diseases, aging, cancer, cell death,
antigen presentation and bacterial invasion [87]. Many neurodegenerative diseases, such as
Alzheimer’s disease, Parkinson’s disease, amyotrophic lateral sclerosis, or Huntington’s disease, are linked to defective autophagy due to the failure to clear aggregates of mutated toxic proteins [88]. In
insects it was shown that increased autophagy is related to the nonapoptotic programmed cell death,
the type II cell death, where increased lysosomal activity and formation of a large number of
autophagosomes can be observed [89]. Such a death pathway was only found in the developing
nervous system of higher eukaryotes [90].
Autophagy is also an adaptive catabolic response to metabolic stresses, including starvation,
growth factor depletion or hypoxia. A key molecule in regulating autophagy is the kinase mammalian
target of rapamycin (mTOR). When mTOR is phosphorylated, autophagy is inhibited until it is
dephosphorylated [86]. mTOR functions as a sensor for the cellular nutrient level. Nutrient deprivation
in cells leads to inhibition of mTOR by dephosphorylation. mTOR inhibition results in activation of the
ULK1/ULK2 (Atg1 in yeast) complex which initiates a cascade of events that lead to the formation of
autophagosomes [91].
Most of the genes involved in autophagy were first studied and characterized in yeast. These
autophagy-related (Atg) genes are involved in regulating autophagosome formation and they also
function in other stages of autophagy. Around thirty-three Atg genes have been identified in yeast and
homologues of many of these genes have been described in higher eukaryotes [92]. Two major steps
are responsible for the process of autophagosome formation: nucleation and elongation of the
isolation membrane (Figure 1.6). Formation of autophagosomes starts with the nucleation process, the
generation of isolation membranes. The nucleation process is activated by phosphatidylinositol (PI)
phosphorylation and involves lipid kinase signalling complex comprising of the ULK1/Atg1 kinase
complex, the autophagy specific class III phosphatidylinositol-3-OH kinase (PI(3)K/hVps34), Beclin 1
-
INTRODUCTION
21
Figure 1.6: The formation of autophagosomes in mammalian cells.
(A) Autophagy is induced by the formation (nucleation) and expansion (elongation) of an isolation membrane (phagophore).
Phagophore closes to form the autophagosome (Aϕ), a double-membrane vesicle that engulfs cytoplasmic components and organelles. Autophagosomes fuse with lysosomes (Ly) to form autolysosomes (AL). (B) The nucleation step is shown involving
the ULK1/Atg1 kinase complex, the autophagy specific PI3-kinase complex (Beclin 1, Vps15, Vps34) and PI(3)P effectors and
their related proteins. Ambra 1 promotes Beclin 1/Vps34 interaction. Additionally, UVRAG and Bif-1 are two regulators of the
Beclin 1/Vps34 complex. (C) Atg12- and LC3/Atg8-conjugation systems are responsible for the elongation step. Atg7 activates
LC3, LC3 is then transferred to Atg3 and modified with a lipid attachment molecule phospatidylethanolamine (PE). The complex
Atg12/Atg5/Atg16L mediates the LC3-PE binding to the autophagosome membrane (adapted from [86]).
(Atg6 or Vps30 in yeast), and the coiled-coil tether p150 (hVps15 or Vps15 in yeast). p150 is important
for the attachment of the complex with the isolation membranes. Phosphatidylinositol-3-phosphate
recruits other effector proteins to the isolation membrane, which in turn promote expansion of the pre-
autophagosomal membrane [84]. There are two major models proposed for the origin of the isolation
membranes, the maturation model and the assembly model [93]. The maturation model states that the
isolation membrane is derived from pre-existing cytoplasmic organelle such as the ER. The assembly
model suggests that the autophagosomal membrane is established anew from localized lipid synthesis
(Figure 1.7) [93].
Elongation of the Isolation membrane is induced by phosphatidylethanolamine (PE)
modification of LC3/Atg8 [86], a light chain of MAP1 (MAP1LC3/LC3) (Figure 1.6) [94]. They were first
identified in neurons and there are three paralogues in mammals: LC3A, LC3B and LC3C [95].
GABARAP, GEC1/GABARAPL1, GATE16/GABARAPL2 and GABARAPL3 are four additional Atg8
homologues [96]. LC3 is cleaved by the cysteine protease Atg4/autophagin to form the cytosolic LC3-
I. HsAtg4B cleaves the LC3 precursor after a conserved glycine residue (Gly120), to which the amino
group of PE is then conjugated [97, 98]. Thus, the modification of LC3-I to membrane-bound form
-
INTRODUCTION
22
LC3-II involves two conserved ubiquitin-like conjugation systems that are arranged in a coordinated
way. First, Atg12 is associated to Atg5, and then the complex formation of LC3-PE, prior to their
conjugation to the isolation membrane. Atg7 acts as an E1-like ubiquitin conjugating enzyme in both
cases, whereas Atg10 and Atg3 function as E2-like ubiquitin conjugating enzymes for Atg12 and LC3-
I, respectively. The Atg12-Atg5 complex is localized together with Atg16 to the isolation membrane
acting as an E3-like ubiquitin conjugating enzyme for the PE attachment to LC3 [99, 100].
Figure 1.7: Models of the origin of the isolation membrane.
The maturation modal states that the membrane may be derived
directly from the ER (A), or in form of vesicular transport (B). The
assembly model states that the membrane may be assembled de
novo at the site of phagophore formation, originating from
nonvesicular transport such as micellar (C) or local synthesis (D)
(adapted from [93]).
After forming the double-membrane vacuole, the autophagosomes undergo extensive
remodelling. This remodelling process, also known as autophagosome maturation, includes fusion
with early and late endosomal vesicles, such as the MVBs, to form the amphisome hybrid organelle.
Finally, amphisomes acquire hydrolytic enzymes through the fusion with lysosomes forming
autolysosomes to degrade intraluminal components and the inner membrane [101, 102]. The digested
components are then recycled to the cytosol and reused. There is a close relation between the
endocytic and autophagic pathway at the molecular level as it was shown that many proteins of the
endosomal pathway are involved in autophagosome maturation steps in yeast and mammalian cells,
such as Rab5 [103], Rab7 [104, 105], Rab11 [106], Vam3p, SKD1 and the SNARE protein Vti1p
(Figure 1.8).
Figure 1.8: Autophagosome and
MVB maturation to autolysosome.
Cytosolic proteins and organelles
destined for degradation are engulfed in
phagophore forming autophagosome.
In the endocytic pathway, proteins
targeted for degradation are sorted into
luminal vesicles of MVBs. Mature MVBs
fuse with autophagosome to form
amphisome. Finally, amphisomes fuse
with lysosome to generate
autophagolysosome (adapted from
[87]).
-
INTRODUCTION
23
1.6 SNAREs
Membrane docking and fusion are fundamental processes in cells with regard to the secretory
and endocytic pathway. One of the key mediators in this processes are the soluble NSF-attachment
protein receptor (SNARE) proteins [3, 107]. SNARE proteins represent a superfamily of the type II
transmembrane proteins, with nine subfamilies, in which totally 36 members were identified in human
[108]. The C-terminal domain of most SNAREs is attached to the membrane, while the N-terminus
points toward the cytosol. The SNAREs possess a single transmembrane domain at their C-terminal
ends that is bound to the SNARE motif by a short linker. The SNARE motif consists of a characteristic,
conserved, and reactive heptad 60-70 residue repeat domain. Between the subgroups of SNAREs,
there are differently folded domains in many SNAREs that are found between the N-terminus and
SNARE motif [109, 110]. SNAREs are grouped according to their domain structure and sequence
homology, as members of the vesicle associated membrane protein (VAMP), syntaxin or
synaptosomal associated protein of 25 kD (SNAP-25) families [111]. Further, SNAREs are classified
according to their localization, into v-SNAREs (vesicle-associated) and t-SNAREs (target membrane-
associated), or, based on a conserved residue in the centre of the SNARE motif, into R- and Q-
SNAREs [112]. The R- and Q-SNAREs terminology is more convincing than the v- and t-SNAREs as
the latter terminology is not persuasive in describing homotypic fusion events, e.g. those between
endosomes. In some respects, there is a rough correspondence of R-SNAREs to v-SNAREs and Q-
SNAREs to t-SNAREs that is only true for fusion events between secretory vesicles and the plasma
membrane. The Q-SNAREs can be further classified into Qa-, Qb- and Qc-SNAREs. The QabcR-rule
states that all functional SNARE complexes contain one member of each subfamily [112].
Membrane fusion starts with the docking of the donor membrane to the target membrane
(Figure 1.9). The first step is the nucleation that describes the formation of a trans-SNARE complex,
consisting of one R-SNARE and three Q-SNAREs. These proteins form an extremely stable four-helix
bundle that brings the lipid bilayers close together and pushes water molecules away from the fusion
site, resulting in membrane fusion. The fusion event is initiated by using the free energy that is
released during the formation of the extraordinary stable four-helix bundle [113, 114]. After membrane
fusion, the trans-SNARE complex becomes the cis-SNARE complex because the SNARE complex
now resides in the same membrane [115]. The SNARE complex needs to be dissociated to allow the
SNAREs to be recycled back to their original location. The recycling of SNAREs is mediated by the
AAA+ protein NSF (N-ethylmaleimide-sensitive factor) and the cofactor α-SNAP. α-SNAP binds to the cis-SNARE complex and recruits NSF, which leads to ATP hydrolysis and to the dissociation of the
complex. The released SNAREs are subsequently recycled back to their resident compartment [116].
Up to date, the neuronal SNARE complex is the best studied among the SNARE complexes.
The R-SNARE VAMP2 resides on the synaptic vesicle membrane, and the two Q-SNAREs syntaxin 1
(Qa) and SNAP-25 (Qbc) are located on the presynaptic plasma membrane. Membrane fusion is
driven by the formation of the stable four-helix bundle that is referred to as the SNARE core complex
[117, 118]. The crystal structure of the neuronal SNARE core complex resolved that the four-helix
bundle consists of one coil of syntaxin 1 and VAMP2, and two coils of SNAP-25 [119]. This coiled-coil
structure is kept together by 15 hydrophobic interaction-planes and a central ionic layer, in which the
arginine and three glutamine of the core motif form strong hydrogen bonds [111]. The neuronal
SNARE complex was first observed in fusion events of synaptic vesicles, with the active zone of
presynaptic membrane. Therefore, the interaction of these three proteins is essential for synaptic
transmission (Squire et al., Fundamental Neuroscience, 2003).
-
INTRODUCTION
24
Figure 1.9: Membrane fusion is mediated by SNAREs.
One monomeric R-SNARE from the donor membrane assembles with the oligomeric Q-SNAREs at the target membrane to
form a four-helix bundle, the trans-SNARE complex. The four-helix bundle initiates membrane fusion leading to the formation of
a cis-SNARE complex at the target membrane. In order to recycle SNAREs, α-SNAP binds to the cis-SNARE complex and recruits NSF. ATP hydrolysis leads to the dissociation of the SNARE complex. The released R-SNARE can bud into vesicles to
recycle back to its resident compartment (adapted from [107]).
Different SNARE complexes are necessary to perform fusion events between distinct cellular
membranes and each SNARE functions in specific intracellular fusion steps that are localized in
different compartments. Syntaxin 1, -2, -4, SNAP-23 and SNAP-25 are localized at the plasma
membrane, VAMP2 is found on synaptic and neurosecretory vesicles, syntaxin 5 and VAMP4 in the
Golgi apparatus, syntaxin 6 at the TGN, syntaxin 7 in late endosomes and lysosomes, and VAMP2/3
and syntaxin 13 in recycling endosomes [107].
In non-polarized cells, syntaxin 13 is localized to sorting endosomes and tubulovesicular
recycling endosomes, in which it is involved in early endosome trafficking and in endosomal recycling
of plasma membrane components [120, 121]. In neurons, syntaxin 13 is also found in tubular sorting
and recycling endosomes, where it colocalizes with transferrin receptor (TfR). Syntaxin 13 is
associated with recycling endosomes that are localized in both dendrites and axons [62].
Neurotransmitter receptor trafficking at dendritic spines from excitatory synapses is regulated by
recycling endosomes containing syntaxin 13 [122-124]. Syntaxin 13 also plays specific roles in neurite
growth [125]. Hence, recycling endosomes may be the source of membrane and protein material for
activity-dependent spine potentiation and growth.
SNAREs ensure specific membrane docking and fusion, but additional specificity is mediated
by tethering proteins that dock the membranes prior to SNARE complex formation. One of the
regulatory proteins determining membrane association is the Rab family of small GTPases. Rab
tethers collaborate with SNAREs to ensure correct fusion between membranes.
-
INTRODUCTION
25
1.7 Rab proteins
Rab GTPases are members of the Ras-like small GTP-binding protein superfamily, which
function as molecular switches that cycle between GTP- (active) and GDP-bound (inactive)
conformations. This family also includes Ras, Ran, ARF and Rac/Rho GTPases. More than 60
different Rab family members were identified in humans and 11 were found in yeast [126, 127].
Different Rab GTPases are localized to the cytosolic face of distinct intracellular membranes, where
they regulate intracellular membrane trafficking [126]. Rab GTPases undergo cycles of membrane
association and cytosolic localization, which is tightly coupled to the guanine nucleotide cycle. This
process is strictly regulated by a large number of Rab associated proteins (Figure 1.10). Rab
GTPases associates with the membrane through a hydrophobic geranylgeranyl tail that is the result of
post-translational modification at the C-terminus. Newly synthesized Rab proteins in the cytosol, in
their GDP-bound state, are recognized and bound by a Rab escort protein (REP), that targets Rab
proteins to Rab prenyltransferase (RGGT) [128]. RGGT prenylates Rab proteins by addition of
geranylgeranyl isoprenoids on one or two C-terminal cysteine residues, present in CC or CXC motifs.
The cytosolic GDP dissociation inhibitor (GDI) forms a complex with GDP-bound Rab proteins, shields
the geranylgeranyl tail from the hydrophilic environment, and therefore maintains Rab soluble in the
cytosol. At the membrane, GDI is then replaced by the GDI displacement factor (GDF) to liberate the
prenyl groups promoting membrane association. Rab proteins are subsequently activated by the
release of GDP, and through the loading of GTP, promoted by guanidine nucleotide exchange factor
(GEF). In the GTP-bound form, Rab is then free to interact with downstream effector proteins. The
active state of Rab proteins is terminated by GTP hydrolysis catalyzed by GTPase activating proteins
(GAPs) [129]. The cycle of membrane association and membrane extraction can then be restarted.
Figure 1.10: Rab GTPase undergoes a
cycle of membrane association and
membrane extraction. Rab forms a complex
with cytosolic GDP-dissociation inhibitor
(RabGDI) after GTP hydrolysis. GDI
displacement factor (GDF) replaces GDI that
promotes association of Rab to the
membrane. GDP is exchanged by GTP
promoted by guanidine nucleotide exchange
factor (RabGEF). In the GTP-bound form,
Rab is then free to interact with downstream
effector proteins. The active state of Rab
proteins is terminated by GTP hydrolysis
catalyzed by GTPase activating proteins
(RabGAP).
In the activated and GTP-bound state, Rab proteins recruit and interact with Rab effector
proteins that represent a very heterogeneous group of proteins. Rab effector proteins cover a variety
of functions. They are involved in the sorting during vesicle formation (e.g. TIP47) [130], in the
regulation of intracellular transport, in which they function as motor or motor adapters (e.g.
Rabkinesin-6, myosinVa) [131, 132], and in vesicle tethering (e.g. EEA1, Golgins, TRAPP-I, TRAP-II,
HOPS, VPS) [133]. Rab proteins also play a role in membrane fusion, where they indirectly regulate
SNARE functions through Rab effectors. Therefore, Rab proteins are crucial in tethering and docking
-
INTRODUCTION
26
processes between organelles, which then lead to membrane fusion. Thus, it has been proposed that
the Rab domain is a prerequisite for the appropriate trans-SNARE pairing upon tethering or docking.
Different Rab GTPases are localized to the cytosolic face of distinct intracellular membranes, where
they regulate endocytic membrane traffic. For example, Rab1 is located at the ER exit sites, and in the
pre-Golgi intermediate compartment (IC), and it is involved in the tethering of ER-derived vesicles to
the Golgi complex. Rab6, Rab33 and Rab40 are localized to the Golgi and mediate intra-Golgi
trafficking. Rab6 also regulates the movement of vesicles and organelles along microtubules by direct
interaction with Rabkinesin-6 [131, 134]. Rab33, together with Rab24, regulate the formation of
autophagosomes. Rab3 is present on the synaptic vesicle surface, and regulates vesicle availability or
docking to release sites. Rab5 is localized to sorting endosomes and plasma membrane, and it
mediates endocytosis and homotypic early endosome fusion of clathrin-coated vesicles [135, 136].
Rab11 is involved in the slow endocytic recycling pathway through recycling endosome, whereas
Rab4 regulates the fast endocytic recycling pathway directly from sorting endosomes. The late
endosome-associated Rab7 plays a role in the maturation of late endosome and their fusion with
lysosomes. Rab9 also resides in late endosomes and mediates trafficking from late endosomes to the
TGN (Figure 1.11) [137].